Dialysis Facilities: Problems Remain in Ensuring Compliance with Medicare Quality Standards

Dialysis Facilities: Problems Remain in Ensuring Compliance with
Medicare Quality Standards (08-OCT-03, GAO-04-63).

Most patients with end-stage renal disease (ESRD) must rely on
dialysis treatments to compensate for kidney failure. Currently,
over 222,000 ESRD patients visit dialysis centers several times a
week to have toxins removed from their bloodstreams. While
dialysis care has improved overall, questions remain regarding
the quality of care provided by some of the nation's roughly
4,000 ESRD facilities. We examined (1) the extent and nature of
quality of care problems identified at dialysis facilities, (2)
the effectiveness of state survey agencies in ensuring that
quality issues are uncovered, corrected, and stay corrected, and
(3) the extent to which the Centers for Medicare & Medicaid
Services (CMS) funds, monitors, and assists state survey
activities related to dialysis care.
-------------------------Indexing Terms-------------------------
REPORTNUM: GAO-04-63
ACCNO: A08680
TITLE: Dialysis Facilities: Problems Remain in Ensuring
Compliance with Medicare Quality Standards
DATE: 10/08/2003
SUBJECT: Diseases
Health care facilities
Health care services
Health surveys
Patient care services
Inspection
Noncompliance
Performance measures
Dialysis
Quality-of-care
Medicare Program

******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO Product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
******************************************************************
GAO-04-63

United States General Accounting Office

GAO

Report to the Chairman, Committee on Finance, U.S. Senate

October 2003

DIALYSIS FACILITIES

Problems Remain in Ensuring Compliance with Medicare Quality Standards

GAO-04-63

Highlights of GAO-04-63, a report to the Chairman, Committee on Finance,
U.S. Senate

GAO suggests that Congress consider authorizing CMS to impose immediate
sanctions, such as monetary penalties or denying payment for new Medicare
patients, on dialysis facilities cited with serious deficiencies in
consecutive surveys. GAO recommends that the CMS Administrator create
incentives for facilities to maintain compliance with quality standards,
increase use of expert staff in conducting ESRD facility surveys, and
enhance the support and monitoring of state survey agencies. CMS did not
indicate an intention to implement five of our six recommendations.

www.gao.gov/cgi-bin/getrpt?GAO-04-63.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact Leslie G. Aronovitz at (312)
220-7600.

October 2003

DIALYSIS FACILITIES

Problems Remain in Ensuring Compliance with Medicare Quality Standards

A substantial number of ESRD facilities do not achieve minimum patient
outcomes specified in clinical practice guidelines, with significant
proportions of their patients receiving inadequate dialysis or treatment
for anemia. Similarly, inspections of dialysis facilities by state survey
agencies have uncovered numerous problems that put patient health at risk.
Between fiscal years 1998 and 2002, these inspections, commonly called
surveys, revealed that 15 percent of facilities surveyed had serious
quality problems that, if left uncorrected, would warrant termination from
the Medicare program. Serious deficiencies commonly found during surveys
included medication errors, contamination of water used for dialysis, and
insufficient physician involvement in patient care.

Infrequent, poorly targeted, and inadequate inspections allow facilities'
quality of care problems to go undetected or remain uncorrected.
Specifically:

o Although ESRD survey activity has increased in recent years, only nine
state survey agencies consistently met CMS's goal to inspect 33 percent of
ESRD facilities annually.

o A substantial number of facilities go many years between inspections.
In fiscal year 2002, 216 facilities nationwide went 9 or more years
without an inspection.

o Deficiencies may not have been detected during an inspection if the
surveyors had little experience in assessing dialysis quality.

Even when deficiencies are identified and facilities take corrective
action, little incentive exists for these facilities to remain in
compliance. Data show a pattern of repeated serious deficiencies in
successive inspections of an individual facility. No effective sanctions
are available to enforce compliance, short of terminating the facility
from the Medicare program, which is rarely done.

Federal monitoring of state agencies' performance of surveys and technical
assistance provided is uneven across CMS regions. CMS substantially
increased its funding for ESRD surveys from an estimated $3.1 million in
fiscal year 1998 to $8.2 million in fiscal year 2002. At the same time,
several CMS regional offices in our study did not actively oversee how the
state agencies used these funds to improve survey activities. CMS has not
taken steps needed to facilitate information sharing between federally
funded ESRD networks and state agencies on the performance of individual
dialysis facilities-information that could help states to target their
inspection resources. In addition, CMS has not offered adequate training
opportunities for surveyors inspecting ESRD facilities.

Contents

Letter

Results in Brief
Background
Quality Problems Prevalent among Dialysis Facilities and Put

Patient Health at Risk
Limitations in the ESRD Survey Process Leave Quality Problems
Undetected or Inadequately Addressed
CMS Has Increased Funding for State Surveys, but Monitoring and

Technical Support Are Uneven
Conclusions
Matter for Congressional Consideration
Recommendations for Executive Action
Agency Comments and Our Evaluation

3 5

24 33 35 35 36

Appendix I Scope and Methodology

Appendix II Medicare Conditions for Coverage for Dialysis Facilities

Appendix III State Agencies' Progress toward Meeting CMS Survey Goals

Appendix IV Comments from the Centers for Medicare & Medicaid Services

Appendix V GAO Contact and Staff Acknowledgments 56

GAO Contact 56
Acknowledgments 56

Related GAO Products

Tables

Table 1: Proportion of ESRD Facilities Recertified Within 3, 6, 9, or More
Years, Fiscal Years 1998 to 2002

Table 2: Association between Surveyor Specialization and Rate of
Condition- and Standard-Level Deficiencies Cited in Fiscal Years 2001 and
2002

Table 3: Rates of Repeated Deficiencies in Consecutive Surveys Conducted
from Fiscal Years 1998 through 2002

Table 4: Federal Support for Provider Surveys, Fiscal Years 1998 to 2001

Table 5: ESRD Facilities Recertified Annually by State, Fiscal Years 1998
to 2002

Table 6: Facilities to Be Recertified to Meet CMS 3-Year Goal, by State 18

20 22 26 45 47

Figures

Figure 1: Projected Growth in the ESRD Population and Medicare Costs 6

Figure 2: Number of Facilities Where Some Patients Receive Inadequate
Dialysis Treatment and Anemia Management, 2000 10

Figure 3: State Variation in the Rate of Condition-Level Deficiencies
Cited in Recertification Surveys Conducted from Fiscal Year 1998 through
2002 12

Figure 4: ESRD Facility Survey Rate Compared to CMS Goal, Fiscal Years
1998 to 2002 16 Figure 5: State Variation in the Proportion of Dialysis
Facilities Surveyed for Recertification, Fiscal Year 2002 17

Abbreviations

CMS Centers for Medicare & Medicaid Services
DFC Dialysis Facility Compare Web site
EPO erythropoietin
ESRD end-stage renal disease
ICF/MR intermediate care facilities for the mentally retarded
LTC long-term care
OSCAR Online Survey Certification and Reporting system

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States General Accounting Office Washington, DC 20548

October 8, 2003

The Honorable Charles E. Grassley
Chairman
Committee on Finance
United States Senate

Dear Mr. Chairman:

Most patients with end-stage renal disease (ESRD)-a life-shortening,
chronic illness-must rely on dialysis treatments to compensate for kidney
failure. Currently, over 222,000 ESRD patients spend 3 to 5 hours at
dialysis centers three times a week, where dialysis machines remove
toxins from their bloodstreams. In addition to having permanent kidney
failure, ESRD patients are likely to suffer from diabetes or heart disease
and are at risk for developing illnesses during their course on dialysis.
Therefore, the care of ESRD patients requires expertise in both the
medical and technical aspects of maintaining patients on dialysis.

While dialysis care has improved overall, according to a 2002 Department
of Health and Human Services report, questions remain regarding the
quality of care provided to Medicare beneficiaries by some of the nation's
roughly 4,000 dialysis facilities. The HHS report noted that many ESRD
patients do not receive treatment meeting the minimum standards
established in the National Kidney Foundation's clinical practice
guidelines, which, when not met, have documented adverse effects on
patient outcomes. In 2001, 16 percent of dialysis patients did not have an
adequate amount of toxins removed from their blood, 24 percent had
anemia that was not brought under control, and 19 percent of patients
were dialyzed for extended periods using catheters, the least effective
and
most risky method for connecting patients to dialysis machines.1

ESRD is the one medical condition that confers eligibility regardless of
age
to the Medicare program, which otherwise pays for health care provided

1Department of Health and Human Services, Centers for Medicare & Medicaid
Services, 2002 Annual Report:EndStageRenal DiseaseClinical Performance
Measures Project (Baltimore, Md.: December 2002). These assessments are
based on the clinical performance measures developed by CMS, building on
the National Kidney Foundation's 1997 Dialysis Outcome Quality Initiative
Clinical Practice Guidelines.

to people who are over 65 years of age or to those with disabilities. The
Centers for Medicare & Medicaid Services (CMS), which oversees the
Medicare program, has responsibility for ensuring that dialysis patients
receive quality care. For this purpose, CMS contracts with state survey
agencies that conduct onsite inspections. Following up on a report we
issued in June 2000,2 you asked us to review CMS's system for enforcing
Medicare's minimum quality and safety standards for ESRD facilities and to
assess whether and how it might be strengthened. Specifically, we examined
(1) the extent and nature of quality of care problems identified at
dialysis facilities, (2) the effectiveness of state survey agencies in
ensuring that quality issues are uncovered, corrected, and stay corrected,
and (3) the extent to which CMS funds, monitors, and assists state survey
activities related to dialysis care.

To address these issues, we obtained data from existing national databases
and original data from 10 states. We analyzed facility-specific
information about quality measures reported on CMS's Dialysis Facility
Compare, a consumer guide available on the Internet. For the nation as a
whole and each of the states,3 we also analyzed data from CMS's Online
Survey Certification and Reporting (OSCAR) system for the last 5 fiscal
years, 1998 through 2002. This database provides information on the dates
when surveys took place, the deficiencies cited, and the time spent
conducting various survey activities. In addition, we interviewed
cognizant officials at CMS's central office and reviewed changes in the
CMS budget devoted to survey activities from fiscal years 1998 to 2002.

To supplement available national data, we obtained additional information
from 10 states-Alabama, California, Florida, Kansas, Maryland,
Mississippi, Missouri, Nevada, New York, and Pennsylvania-which together
accounted for more than one-third of all facilities in fiscal year 2001.
They were selected to provide variation across a range of dimensions,
including the proportion of ESRD facilities surveyed and deficiencies
cited, number of ESRD facilities, and geographic diversity. We interviewed
state surveyors and administrators, representatives from

2U.S. General Accounting Office, MedicareQuality of Care: Oversight of
Kidney Dialysis Facilities Needs Improvement, GAO/HEHS-00-114 (Washington,
D.C.: June 23, 2000). This report highlighted the need for additional
enforcement tools to ensure that corrections of quality problems
identified in surveys of ESRD facilities would be sustained over time. It
also urged improved cooperation and data sharing between state survey
agencies and ESRD networks to improve targeting of facilities selected for
inspection.

3In this report, "states" refers to the 50 states and the District of
Columbia.

Results in Brief

ESRD networks (organizations that promote quality improvement in ESRD
services), and federal regional office officials responsible for
monitoring ESRD facility surveys. In addition, we collected detailed
information on several states' corps of ESRD surveyors, including their
background, training, and experience. We also examined the written reports
from numerous facility surveys conducted within the last 2 years. (App. I
contains more detail on our scope and methodology.) Our work was conducted
from August 2002 to September 2003 in accordance with generally accepted
government auditing standards.

A substantial number of dialysis facilities do not achieve the minimum
patient outcomes specified in clinical practice guidelines for a
significant proportion of their patients. Data reported on Dialysis
Facility Compare show that, in 2000, 512 facilities had 20 percent or more
of their patients receiving inadequate dialysis treatment, and nearly
1,700 facilities had 20 percent or more of their patients receiving
inadequate care for anemia. In addition, the CMS-funded system of on-site
inspections of facility conditions, equipment, and staffing has uncovered
numerous problems that put patient health at risk. From fiscal year 1998
through 2002, these inspections, generally called surveys, revealed that
15 percent of facility surveys identified serious quality problems that,
if left uncorrected, would warrant termination from the Medicare program.
Serious deficiencies commonly found during surveys included medication
errors, contamination of water used for dialysis, and insufficient
physician involvement in patient care.

Infrequent, poorly targeted, and inadequate inspections by state survey
agencies allow facilities' quality of care problems to go undetected or
remain uncorrected. Specifically:

o Although ESRD survey activity has increased in recent years, state
compliance with CMS's goal to resurvey 33 percent of ESRD facilities
annually has been inconsistent. While 33 states met the goal in at least 1
of the last 2 fiscal years, only 9 of the 33 states surveyed a third or
more of their facilities in both years. Eighteen states failed to meet the
goal in either fiscal year 2001 or 2002.

o A substantial number of facilities go many years between inspections.
In fiscal year 2002, 216 facilities nationwide (5.4 percent) went 9 or
more years without an inspection, up from 53 facilities (1.6 percent) in
fiscal year 1998.

o Deficiencies may not have been detected during a survey if the
surveyors who inspected the facilities had little experience in assessing
dialysis

quality. Data from several states showed that survey agencies where
designated staff specialized in performing ESRD surveys uncovered a
substantially larger number of deficiencies than agencies without such
staff expertise.

Even when deficiencies are identified and facilities take corrective
action, little incentive exists for these facilities to remain in
compliance with Medicare's minimum quality standards on a continuing
basis. As shown in nationwide data, when quality problems were cited, the
problems were corrected but often did not stay corrected. For example,
from fiscal years 1998 through 2002, 18 percent of facilities found to
have serious deficiencies were cited again for the same deficiencies in
successive inspections. At present, there is no effective sanction to
encourage a facility to avoid repeating prior deficiencies, short of
terminating the facility from the Medicare program, which is rarely done.

CMS has expanded funding to support state ESRD survey activities, but its
monitoring of state agencies' performance of surveys and providing
technical assistance is uneven across CMS regions. CMS substantially
increased its aggregate funding for ESRD surveys from an estimated $3.1
million in fiscal year 1998 to $8.2 million in fiscal year 2002. At the
same time, several regional offices in our study did not actively oversee
or assist in improving ESRD survey activities. In addition, CMS has not
removed barriers between federally funded ESRD networks and state agencies
that inhibit the sharing of information on the performance of individual
dialysis facilities-information that could assist states in targeting
their inspection resources. Furthermore, surveyors in several states
reported that CMS has not offered adequate training opportunities for
surveyors inspecting ESRD facilities.

To encourage ESRD facilities to adhere to Medicare quality standards, we
suggest that Congress consider authorizing CMS to impose immediate
sanctions, such as monetary penalties or denying payment for new Medicare
patients, on dialysis facilities cited with serious deficiencies in
consecutive surveys. We are also recommending that CMS: conduct more
frequent surveys of facilities with serious deficiencies; publicize
facilities' survey results; encourage state agencies to use
ESRD-specialized surveyors; expand ESRD surveyor training opportunities;
require periodic, routine sharing of information between ESRD networks and
state survey agencies; and enhance oversight of state agency performance.

Background

In its comments on a draft of this report, CMS affirmed its commitment to
strengthening oversight of dialysis facilities and state survey agencies,
but did not indicate an intention to implement five of our six
recommendations. Instead, the agency highlighted its efforts to develop
tools to assist states in selecting facilities for inspection and to make
the survey process more uniform. We continue to believe that more focused
efforts to evaluate compliance with Medicare requirements and stronger
actions against poor performers are needed to ensure an effective,
consistent, and timely ESRD survey and certification program.

Individuals with ESRD, characterized by permanent kidney failure, must
undergo either regular dialysis treatment or a kidney transplant to stay
alive. In 2000, about 248,000 individuals received one of two modes of
dialysis treatment-hemodialysis or peritoneal dialysis-both of which can
be performed at a facility or at home.4 Most ESRD patients undergo
hemodialysis.5 The number of hemodialysis patients enrolled in Medicare
has risen sharply, from about 118,000 in 1991 to over 222,000 in 2000.
With anticipated annual growth of over 7 percent, the dialysis population
is projected to reach more than 520,000 by 2010.6 (See fig. 1.) This
growth in enrollment has been attributed largely to improvements in the
survival rate for people with ESRD and an increase in the number of
Americans with conditions, such as diabetes or high blood pressure, that
often lead to kidney failure.

4In hemodialysis, a patient's blood is filtered through an external
machine that acts as an artificial kidney to withdraw excess fluids and
toxic materials before returning cleansed blood to the patient. The
machine uses a semipermeable membrane, called a hemodialyzer, to filter
out the toxins. In peritoneal dialysis, the patient's peritoneal membrane,
located within the abdominal cavity, is used to remove excess fluids and
toxins.

5In 2000, about 222,300 patients received hemodialysis, 21,400 underwent
peritoneal dialysis, and 4,400 underwent dialysis of an unspecified mode.
In addition, approximately 74,700 beneficiaries were recipients of kidney
transplants, for a total of approximately 322,800 individuals that
received Medicare benefits as of December 31, 2000.

6Projections are based on data for 1982 to 1997. See J.L. Xue, J.Z. Ma,
T.A. Louis, and A.J. Collins, "Forecast of the Number of Patients With
End-Stage Renal Disease in the United States to the Year 2010," Journal of
theAmerican Society ofNephrology, vol. 12 (2001): 2753-2758.

Figure 1: Projected Growth in the ESRD Population and Medicare Costs

Growth in the ESRD population has been matched by growth in the number of
dialysis facilities. In the decade between 1991 and 2001, the number of
outpatient dialysis facilities doubled from about 2,000 to more than 4,000
facilities. In 2001, 83 percent of all facilities were freestanding
(nonhospital-based) and 79 percent of all facilities were for-profit. In
2001, the four largest for-profit dialysis chains accounted for about
two-thirds of all freestanding facilities.

The rise in the ESRD population has been accompanied by an even more rapid
increase in program spending. Medicare not only provides coverage

to most beneficiaries with ESRD for all ESRD-related services but for
their other health care needs as well.7 From 1990 to 2001, Medicare
expenditures for beneficiaries with ESRD rose from about $5 billion to
over $15 billion, and are forecast to grow to $28 billion in 2010.
Spending growth has been fueled by an expansion of enrollees with greater
medical needs-older beneficiaries and those with chronic
comorbidities8-and the program's inclusion of new treatments, particularly
erythropoietin (EPO)-a synthetic hormone widely used to manage anemia-and
other injectable medications. While Medicare pays ESRD providers a set
amount-a composite rate-including the nursing services provided and
supplies used in each dialysis treatment, it pays separately for
injectable drugs.9 The composite rate for dialysis services has remained
virtually unchanged since the program's inception. However, payments to
freestanding dialysis facilities for injectable drugs have grown
considerably in recent years, increasing from 33 percent of total payments
in 1997 to 40 percent in 2001.

In 1976, CMS established minimum requirements that dialysis facilities
must meet in order to receive Medicare payments. The regulations, referred
to as "conditions for coverage," address 11 general areas, including the
facility's physical environment and overall management by a governing
body, as well as the adequacy of patient treatment plans.10 (See app. II.)
One condition covers the detailed procedures that facilities must follow
if they choose to reuse certain supplies, such as dialyzers, rather than
replace them for each treatment.11 Under each condition are related
"standards." For example, under the condition "physical environment,"

7For individuals eligible for Medicare only because of permanent kidney
failure, Medicare coverage starts on the fourth month of dialysis.
Medicare will not pay for services during the first 3 months of dialysis
unless the patient already has Medicare because of age or disability.
After that, Medicare is the secondary payer for 30 months. During this
period, private insurance or Medicaid pays first on health care bills and
Medicare pays second. Full Medicare coverage begins with the 34th month of
dialysis and any private insurer becomes the secondary payer. For those
who are uninsured, Medicare is the primary payer.

8The proportion of new ESRD patients 75 or older grew from 18 percent in
1991 to about 25 percent in 2001, while the proportion of new ESRD
patients with diabetes grew from 36 percent of all new patients to 46
percent during the same period.

9In 2002, the average composite rate was approximately $130 for
freestanding dialysis facilities. Payments for injectable drugs averaged
about $80 per treatment in 2001.

10See42 C.F.R. Part 405 Subpart U (2002).

11These requirements include appropriate methods for disinfection and
steps to ensure that such supplies are only reused by the same patient.

there are specific standards to maintain the purity of water used for
dialysis. Even deficiencies found solely at the standard level indicate
potential harm to patients. But, deficiencies cited at the condition level
are the most egregious, as they indicate a problem that is widespread at a
facility or serious in terms of its harm, or potential to harm patients.
Typically, they are accompanied by multiple standard-level deficiencies
under that condition.

To ensure provider compliance with dialysis quality standards, Medicare
contracts with state survey agencies.12 These agencies conduct initial
on-site surveys of dialysis facilities when providers seek enrollment in
the Medicare program. Subsequently, state agencies periodically conduct
unannounced inspections, referred to as recertification surveys, to ensure
that facilities are maintaining compliance with Medicare standards.
Although no statutory or regulatory requirements exist regarding the
frequency of recertification surveys, CMS has established goals for state
survey agencies to ensure that facilities are surveyed within certain
intervals. States are expected to survey 33 percent of their dialysis
facilities annually, and each facility every 3 years. In addition, state
survey agencies must respond to complaints that they receive concerning
dialysis facilities and, when warranted, conduct on-site investigations.

If the state agency determines that a facility is out of compliance with
any condition or standard, CMS requires that the facility develop a plan
to correct the deficiency. The state agency is then responsible for
determining if the plan of correction is adequate to address the quality
problems identified. Facilities that do not correct condition-level
deficiencies within a reasonable amount of time, generally within 90 days,
are subject to termination from the program. A much shorter time frame for
termination applies in situations where a facility's noncompliance poses
an immediate and serious threat to patient health or safety.

CMS also contracts with 18 ESRD network organizations that are responsible
for helping providers improve the quality of care patients receive in
dialysis facilities. Rather than enforcing compliance with federal quality
regulations, the networks recruit facility participation in national and
regional quality improvement projects that focus on enhancing

12These agencies are typically part of state health departments and are
responsible for monitoring compliance with quality standards associated
with several types of facilities, including nursing homes and home health
agencies.

specific clinical outcomes of dialysis patients. Networks collect data
from individual facilities on numerous clinical indicators and provide
them feedback on their performance. The networks also provide technical
assistance to facilities and handle grievances concerning patient care.
Each network has a medical review board composed of dialysis facility
representatives, physicians, and dialysis patients, that oversees network
operations.

To assist beneficiaries with ESRD in deciding where to get dialysis
services, CMS reports certain information on Dialysis Facility Compare, an
Internet Web site. Initiated in 2001, the site provides information on
specific characteristics-such as the location, operating hours, and size-
of all Medicare-certified facilities. It also provides data on clinical
outcomes related to several quality measures, but does not contain the
results of state agency surveys. In contrast, CMS routinely posts survey
results for nursing homes on a similar but separate Internet Web site
called Nursing Home Compare.

Data made public by CMS reveals that poor care is a problem at many
facilities, with large numbers of patients receiving inadequate
hemodialyis or treatment for anemia. Similarly, inspections of ESRD
facilities continue to find evidence that serious health and safety
problems exist for dialysis patients. From fiscal year 1998 through 2002,
as many as one out of seven surveys identified problems sufficiently
severe to initiate the process of terminating the facility from the
Medicare program. These deficiencies, such as medication errors and
contamination of water used for dialysis, put the health of patients at
risk.

Quality Problems Prevalent among Dialysis Facilities and Put Patient
Health at Risk

Many Facilities Do Not Provide Adequate Care to Their Hemodialysis
Patients

Data reported on the Dialysis Facility Compare Web site provides evidence
that the care delivered at many facilities is substandard. The most recent
information available indicates that, in 2000, a substantial number of
facilities did not provide all of their Medicare patients with a level of
care that meets minimum clinical practice guidelines. Figure 2 shows the
extent to which facilities did not achieve two commonly accepted quality
benchmarks based on the National Kidney Foundation guidelines: (1) the
percent of the facility's patients not receiving adequate hemodialysis and
(2) the percent of the facility's patients receiving EPO whose anemia was
not adequately managed.13 Despite some measurement limitations, both of

13EPO is used for the treatment of anemia for nearly all dialysis
patients.

these indicators are considered characteristics of patient care that
reflect dialysis facility quality.

Figure 2: Number of Facilities Where Some Patients Receive Inadequate
Dialysis Treatment and Anemia Management, 2000

1000

Number of facilities

800

600

400

200 0

Less than 5 5 to 9 10 to14 15 to19 20 to 24 25 or more Percentage of the
facility's patients with inadequate hemodialysis

1200 Number of facilities

1000

800

600

400

200

0 Less than 10 10 to 19 20 to 29 30 to 39 40 to 49 50 or more

Percentage of the facility's patients with inadequate anemia control

Source: CMS, Dialysis Facility Compare Web site.

Notes: Adequacy of dialysis is measured as the percentage of the
facility's hemodialysis patients that had the minimum recommended urea
reduction ratio-a measure of the waste products removed from the blood-of
65 or more. Data were reported for 3,158 facilities.

Anemia management is measured as the percentage of the facility's patients
who received EPO that had a hematocrit level-a measure of low red blood
count-of 33 or greater. Data were reported for 3,325 facilities.

Relatively few dialysis facilities reported meeting these two national
guidelines for 100 percent of their patients. At about half of the
facilities, fewer than 10 percent of their patients fell short of the
hemodialysis guideline, but at 512 facilities, 20 percent or more of their
patients received inadequate hemodialysis. Results for anemia treatment
were less favorable overall. Nearly 1,700 facilities fell short of meeting
the guideline for anemia management for 20 percent or more of the patients
in their care; at 135 facilities, more than 50 percent of patients
received inadequate treatment for anemia. Research has shown that
variation in such patient outcomes as dialysis adequacy is largely
attributable to factors at the facility-its policies governing dialysis
care, associated practice patterns, and attention to individual patient
problems-as opposed to patient-specific causes.14

Facility Inspections Identify an Unacceptable Level of Serious Quality
Problems

The cumulative results of surveys conducted from fiscal years 1998 through
2002 suggest that condition-level deficiencies-quality problems severe
enough to warrant termination from the Medicare program unless corrected
within 90 days-are still far from rare. Fifteen percent of recertification
surveys conducted nationwide from fiscal year 1998 through 2002 reported
one or more condition-level deficiencies. The distribution across states
of condition-level deficiencies cited was substantially uneven. Several
states reported no condition-level deficiencies during that 5-year period,
whereas other states found such deficiencies in roughly 60 percent of
their surveys. As shown in figure 3, most states were at the lower end of
the range, with 39 states citing condition-level deficiencies in fewer
than 20 percent of their surveys, and 21 states, in fewer than 10 percent
of their surveys.

14J.C. Fink, S.A. Blahut, A.E. Briglia, and others, "Effect of
Center-Versus Patient-Specific Factors on Variation in Dialysis Adequacy,"
Journal ofthe American Society ofNephrology, vol. 12 (2001): 164-169.

Figure 3: State Variation in the Rate of Condition-Level Deficiencies
Cited in Recertification Surveys Conducted from Fiscal Year 1998 through
2002

20 Number of states

Our review of recertification survey reports from fiscal years 2001 and
2002, collected from the 10 states in our study, identified
condition-level deficiencies that were commonly cited among noncompliant
facilities. Multiple instances were found of inadequate clinical
management, medication errors, improper use of reusable dialysis
equipment, contamination of water used for dialysis, and insufficient
professional medical involvement in the dialysis patients' care. State
surveyors documented these problems after reviewing facility personnel
files, policies, procedures, and the facility's overall environment. In
addition, surveyors reviewed a random sample of medical records from 10
percent of the facility's patients.15 The vignettes presented below-which
illustrate the types of problems found in 35 percent of all surveys
conducted from fiscal year 1998 through 2002-were extracted from
surveyors' findings

15A patient's medical record contains required information on identified
problems, a plan of care, and documentation tracking the treatments
actually provided. The record must show ongoing assessments of patient
needs as well as evidence that patients participate in developing their
treatment plans and are informed of outcomes.

10 8

0 0 1 to 9 10 to 19 20 to 29 30 or more

Percentage of the state's surveys that found condition-level deficiencies

Source: GAO analysis of CMS OSCAR data.

Problems Cited at ESRD Facilities Create the Potential for Harm to
Patients

reports. Registered nurses with substantial ESRD survey experience, who we
asked to comment on the clinical implications of these findings, indicated
that the deficiencies could lead in some cases to severely adverse patient
outcomes.

o Failure to monitor laboratory values and medication supply. A Maryland
surveyor found that for 31 days, one facility did not provide any of its
patients with EPO, a medication routinely used to stimulate the production
of red blood cells that are compromised by chronic kidney disease. Upon
reviewing patients' medical records, 8 out of 10 sampled records indicated
that the patient's red blood cell count was below normal, thus requiring
EPO. In addition, 5 of these records showed that the patient's red blood
cell level decreased over a 4-month period. The facility's head nurse did
not monitor and report the patients' abnormal laboratory values to the
physicians and did not respond to the patients' complaints of feeling
tired and lacking energy.

According to our nurse reviewers, patients who have a diminished red blood
cell count for an extended period of time can develop health-related
complications, including heart irregularities and a decrease in brain
function.

o Failure to administermedication as prescribed. A California surveyor
cited a condition-level deficiency when she found that physician orders
were not being followed. One patient's medical record documented that
6,000 units of EPO were prescribed for each dialysis treatment but that
the patient received only 600 units at each treatment for 20 treatments.
Staff confirmed that the patient was receiving the wrong dose, and when
questioned by the surveyors, could not provide an explanation. Another
patient's medical record revealed that, despite a physician-ordered
increase in EPO, the patient received an incorrect dosage of the
medication for almost 2 months. Again, staff acknowledged that the order
to increase the dosage was not carried out. A review of two more patients'
medical records showed written orders for Venofer, a medication to treat
iron deficiency. The records documented that both patients failed to
receive this medication for a week or more. Staff acknowledged that there
was a period of time during which the facility ran out of the medication.

Our nurse reviewers reported that a reduction of Venofer or EPO could
increase the dialysis patients' risk for anemia, a condition that, as
noted above, can cause a patient to experience extreme fatigue and
eventually clinical impairments to the heart and brain.

o Failure to administer dialysis treatments as prescribed. A
recertification survey in Pennsylvania discovered that, for over half of
the medical records reviewed, the facility did not ensure that diagnostic
and therapeutic orders were followed. Specifically, documentation in
patients' medical records revealed that the duration of dialysis
treatments deviated from the amount of time prescribed by a physician. One
patient's medical record indicated that dialysis treatments were ordered
for 3.5 hours in duration. However, actual treatment periods were all less
than the prescribed amount-by 20 to 90 minutes. Similarly, another
patient's record indicated that dialysis treatments were ordered for a
duration of 3 hours and 45 minutes but most treatments were for shorter
duration-as much as an hour less.

Nurse reviewers indicated that when the dialysis treatment period is
reduced, the patient retains toxins and other fluids that have not been
removed adequately from the blood stream. This condition can adversely
affect the patient's overall general health and lead to loss of appetite,
swelling, fatigue, shortness of breath, and possibly heart failure.

o Failure to monitor concentration of chemicals in the water system. A
New York surveyor found that a facility did not monitor the purity of
water used for dialysis. The water used to prepare dialysate, a solution
that removes wastes from the blood during dialysis, contained chemical
contaminates in excess of allowed concentrations. For at least 8 months,
fluoride levels were 1.0-five times greater than the maximum allowable
limit of 0.2. In addition, two water tests showed that calcium levels were
above 5.25, well above the maximum allowable limit for calcium of 2.0. The
facility medical director did not monitor the results of water tests
conducted and did not ensure that the facility's staff took appropriate
action, such as reporting abnormal values or resampling the water.

Nurse reviewers told us that excessive amounts of fluoride could cause a
dialysis patient's red blood cells to rupture and clot and that excessive
amounts of calcium in the blood could increase the incidence of bone
disease.

o Failure to involvea transplant surgeon in the review of patients'
long-term care plans. A recertification survey in Mississippi revealed
that the facility did not involve a transplant surgeon, as required, in
the review of patients' long-term care plans. All of the medical records
reviewed in that facility had long-term care plans that were not updated
within the required 6-month time frame. The surveyor interview with the
facility's medical director confirmed that a transplant surgeon or his
designee had not examined patients' long-term care plans.

Limitations in the ESRD Survey Process Leave Quality Problems Undetected
or Inadequately Addressed

Nurse reviewers commented that, until screened by a transplant surgeon,
the dialysis patient's potential for kidney transplantation cannot be
properly assessed.

Infrequent or poorly targeted inspections allow facilities' quality of
care problems to go undetected or remain uncorrected. Although state
survey activity increased from fiscal year 1998 to 2002, numerous state
agencies did not meet the goal currently set by CMS to survey 33 percent
of all ESRD facilities annually. An increasing number of facilities
continued to operate 9 or more years between inspections. In addition,
states that relied primarily on surveyors with limited experience in
conducting inspections of ESRD facilities tended to report substantially
fewer deficiencies than states using more experienced surveyors,
suggesting that surveyors in the first group of states may have missed
some quality problems. We also found patterns of repeated condition-level
deficiencies, and particularly, citations for the same problem in
successive inspections of an individual facility. Finally, facilities had
little incentive to ensure continued adherence to Medicare's minimum
quality standards in the absence of sanctions for noncompliance other than
termination from the Medicare program-which, historically, has been rarely
used.

Increased CMS Goals Have Led to Greater Survey Activity, but Many States
Fall Short

In recent years, CMS has underscored the importance of conducting
recertification surveys of ESRD facilities by raising its expectations for
the state agencies regarding the frequency with which such surveys should
take place. In fiscal year 2001, CMS increased the recertification goal
for states to 33 percent of facilities each year, up from 10 percent in
fiscal year 1999 and 17 percent in fiscal year 2000. Moreover, since
fiscal year 2001, there has been a parallel goal for states to survey
every dialysis facility within a 3-year period. Thus, by the end of fiscal
year 2003, no dialysis facility should have gone more than 3 years since
its last recertification survey.

In response to CMS's heightened expectations, state agencies surveyed more
ESRD facilities, but not enough to fully meet CMS's current goals. As
shown in figure 4, the percentage of ESRD facilities undergoing
recertification surveys annually grew substantially from fiscal year 1998
to 2001. However, collectively, state agencies did not achieve the current
goal, effective in 2001, of surveying 33 percent of all ESRD facilities
each year. In fact, after increasing to over 28 percent in fiscal year
2001, the survey frequency rate declined to about 27 percent in fiscal
year 2002.

Figure 4: ESRD Facility Survey Rate Compared to CMS Goal, Fiscal Years
1998 to 2002

Percent

35 33 33

0 1998 1999 2000 2001 2002

CMS annual survey goal

Facilities surveyed

Source: GAO analysis of CMS OSCAR data.

Underlying this aggregate trend are wide disparities in survey frequency
rates across the individual state agencies, as shown in figure 5. State
recertification survey rates ranged from zero to 89 percent in fiscal year
2002. Even among the 13 states with the largest number of ESRD
facilities,16 recertification survey rates varied widely-from 10 percent
to 40 percent.

16This top quartile of states represents 60 percent of all ESRD facilities
and 64 percent of all dialysis patients.

Figure 5: State Variation in the Proportion of Dialysis Facilities
Surveyed for Recertification, Fiscal Year 2002

0 0 to 4 5 to 9 10 to 14 15 to 19 20 to 24 25 to 29 30 to 34 35 to 39 40
or more

Percentage of the state's facilities surveyed

Source: GAO analysis of CMS OSCAR data.

While 33 state survey agencies met the expanded CMS survey frequency goal
in at least 1 of the last 2 fiscal years-sometimes by substantial
margins-only 9 of those states met the 33 percent goal in both years. (See
table 5 in app. III.) By contrast, 18 state agencies failed to reach 33
percent in either of the two most recent fiscal years, including some of
the largest ESRD states, such as California, Michigan, Pennsylvania, and
Virginia.

As a result, many states may have difficulty meeting CMS's second goal for
state recertification activity, to survey all their ESRD facilities within
a 3-year period. Because this goal was established in fiscal year 2001,
the first test of state compliance will come at the end of fiscal year
2003. Based on the facilities surveyed in fiscal year 2001 and 2002, 35
states will have to inspect more than a third of their ESRD facilities in
fiscal year 2003 if they are to meet the 3-year goal. (See table 6 in app.
III.) About one in five states has more than 60 percent of facilities left
to survey. Alabama has the most facilities-89 percent-that need to be
surveyed in the current fiscal year. Among the largest states, California
and Virginia have the largest backlogs to overcome-around 76 percent.

Despite improvement in the overall rate of ESRD facility surveys, a
significant proportion of dialysis facilities continue to operate for long
periods without inspections. For example, as of September 30, 2002, 466
facilities had not been surveyed for 6 or more years, of which 216 had not
been inspected for recertification in 9 or more years. Most of the effort
to shorten the interval between recertification surveys has focused on

reducing the number of facilities surveyed within 3 to 6 years. (See table
1.) From fiscal year 1998 to 2000, the proportion of facilities not
surveyed for more than 6 years rose sharply (from 9.8 to 17.4 percent) and
then declined (to 11.6 percent). Those that operated 9 or more years
without a recertification survey steadily increased from 1.6 percent (53
facilities) in fiscal year 1998 to 5.4 percent (216 facilities) in fiscal
year 2002. This aggregate result reflected highly variable survey rates
across states. Four states-California, Texas, New York, and
Missouri-accounted for 174 facilities that had not been surveyed within 9
years by the end of fiscal year 2002.

Table 1: Proportion of ESRD Facilities Recertified Within 3, 6, 9, or More
Years, Fiscal Years 1998 to 2002

Percentage of facilities subject to a recertification survey

Length of time since
last recertification 1998 1999 2000 2001 2002
survey (n=3,250) (n=3,462) (n=3,679) (n=3,882) (n=4,011)

Less than 3 years 51.6 51.2 49.8 62.5 72.4
3 to <6 years 38.6 32.8 32.8 22.9 16.0
6 to <9 years 8.2 13.9 14.2 10.1 6.2
9 or more years 1.6 2.1 3.2 4.4 5.4

Source: GAO analysis of CMS OSCAR data.

State agencies have to balance their efforts to meet survey workload goals
for ESRD facilities against the demands on inspection staff to meet other
CMS survey requirements. In particular, state agencies are required to
inspect nursing homes every 15 months,17 intermediate care facilities for
the mentally retarded (ICF/MR) at least annually, and home health agencies
at least once every 3 years. In its letter to state agencies on fiscal
year 2003 program requirements and budget guidelines for survey
activities, CMS made inspections of dialysis facilities and nine other
types of providers lower in inspection priority, behind nursing homes,
ICF/MRs, and home health agencies.18 ESRD recertifications also received
lower

17The statewide average interval between standard surveys must be 12
months or less. See 42 C.F.R. S: 488.308(b).

18Department of Health and Human Services, CMS, memorandum from the
Director, Survey and Certification Group, "Fiscal Year (FY) 2003 State
Survey and Certification Budget Call Letter - ACTION," July 2, 2002.

priority than investigation of complaints filed against all types of
providers.19 CMS officials asserted that they provide the state survey
agencies with sufficient resources to fulfill expectations across all
provider types. Nonetheless, several state officials we spoke with
reported difficulty in meeting all of these expectations, especially those
experiencing substantial growth in ESRD facilities in their states. They
indicated that, given the relatively low priority assigned to ESRD
recertifications, they would most likely cope by adjusting the number of
dialysis facilities inspected.

Lack of Surveyor Specialization May Contribute to Less Effective Surveys

Even when facilities are inspected, some surveyors may be more adept than
others at identifying quality problems. Because dialysis treatment is
technically complex, surveyors who focus on ESRD surveys say that they
become more proficient in detecting and properly documenting quality of
care problems as a result. However, state agencies may be reluctant to
designate a subset of surveyors who specialize in performing ESRD
inspections as it limits their flexibility in scheduling inspections of
nursing homes, home health agencies, and other provider types. Moreover,
such specialization is less feasible for states with few ESRD facilities
overall. In states without a specialist approach to facility inspections,
many surveyors are likely to conduct no more than a few ESRD surveys each
year. Among the nine state survey agencies from which we collected
workload data, six typically assigned ESRD inspections to surveyors who
spent most of their time surveying other provider types.20 The other three
assigned most ESRD inspections to surveyors who often performed surveys of
dialysis facilities.

19CMS requires every state to establish a screening mechanism to evaluate
complaints as they come in, and to apply explicit criteria to determine
which ones need to be followed up with a survey as well as the time frame
within which that survey must take place. Surveys prompted by complaints
are intended to address a particular issue raised in the complaint, which
often does not involve clinical issues. If during the course of the
complaint investigation the surveyor discovers systemic quality problems,
the inspection is usually converted into a recertification survey.

20For each state, we calculated a specialization ratio that indicated the
likelihood that any given ESRD survey would be conducted by a surveyor who
frequently conducted surveys of dialysis facilities. (See app. I.) On a
scale of zero to one, the values of the specialization ratio clustered
into two groupings: the states with specialized ESRD staff included New
York (0.68), California (0.63), and Maryland (0.57); the states without
ESRD specialized staff included Pennsylvania (0.36), Missouri (0.27),
Alabama (0.21), Florida (0.17), Kansas (0.14), and Nevada (0.11).

A comparison of survey results between states that had a designated corps
of ESRD surveyors and those that did not suggested that surveyors who
frequently conduct ESRD inspections may be more effective in detecting and
reporting deficiencies. Table 2 shows that the more specialized group of
states was almost three times as likely to find a condition-level
deficiency. Surveyors from these states cited a substantially larger
number of deficiencies at the less serious "standard-level" as well. While
other factors could have also influenced the number of deficiencies
reported by surveyors in various states, the magnitude of the difference
observed between states that did and did not specialize suggests that
specialization has a major impact.21

Table 2: Association between Surveyor Specialization and Rate of
Condition- and Standard-Level Deficiencies Cited in Fiscal Years 2001 and
2002

State surveyor specialization in ESRDa

Percentage of Percentage of surveys in surveys in nonspecialized states
specialized states (n=367) (n=261)

Surveys with condition-level
deficiencies 8.4 24.1

Surveys with standard-level deficiencies numbering

26.4 6.9

1 to 5 45.8 21.1
6 to 10 14.2 31.0
11 to 20 11.4 28.4
21 or more 2.2 12.6

Source: GAO analysis of state-provided workload data and CMS OSCAR data.

aNonspecialized states include Pennsylvania, Missouri, Alabama, Florida,
Kansas, and Nevada. Specialized states include, California, Maryland, and
New York.

The importance of surveyor specialization for inspection results may be
stronger for ESRD facilities than other types of providers. Although some
general surveying skills apply across provider types, much of the content
of ESRD standards is highly specialized, reflecting both the technological

21Statistical tests (chi square) indicate that the difference in outcomes
between these two groupings of states is highly significant (p=0.000).
Thus, it is very unlikely that these differences could have occurred
simply by chance.

complexity of the dialysis process and the clinical complexity and
vulnerability of the ESRD patient population. In a 184-page appendix
devoted to ESRD surveys, CMS's State Operations Manual lays out the
specific steps that surveyors are expected to follow. Presumably,
surveyors who have the opportunity to focus on mastering this material
develop greater proficiency in identifying quality of care problems,
including proficiency in identifying indications of adverse patient
outcomes and appropriate facility responses.

Facilities with Prior Deficiencies Are Likely to Be Cited for Problems in
Subsequent Surveys

Our June 2000 ESRD report described the inability of Medicare's survey and
certification system to ensure that problems identified in surveys and
addressed by a facility's plan of correction will stay corrected for the
long term. Once a facility has been recertified, it faces no adverse
consequences should it fail to remain in compliance in the future. When
the next survey takes place-usually several years later-the process will
start over with deficiencies identified and a new opportunity for the
facility to correct them. This allows facilities to cycle in and out of
compliance with Medicare's quality standards.

The results of surveys conducted from fiscal year 1998 through 2002 showed
that a pattern of persistent noncompliance with quality standards was not
uncommon. First, facilities cited for deficiencies in previous surveys
were substantially more likely than other facilities to have deficiencies
when surveyed again. Of surveys involving facilities that had a
condition-level deficiency in their most recent prior survey, 29 percent
had a condition-level deficiency in the subsequent survey as well,
compared with 16 percent for those with only standard-level deficiencies
in the prior survey and 12 percent for those with no prior deficiencies.

Similarly, we found that repeated citations for the same deficiency
occurred frequently. From fiscal year 1998 through 2002, 2,073
recertification surveys (57 percent of the total) involved facilities that
had received deficiencies in their most recent prior survey. Of those, a
third found deficiencies that repeated one or more specific condition- or
standard-level deficiency codes cited in that prior survey. Moreover, 18
percent of the facilities with a condition-level deficiency on the prior
survey were cited again for the same condition-level deficiency. (See
table 3.) Another 44 percent repeated one or more standard-level
deficiencies.

Table 3: Rates of Repeated Deficiencies in Consecutive Surveys Conducted
from Fiscal Years 1998 through 2002

Percentage of subsequent surveys identifying the same deficiencies

Condition-level

Standard-level only

Percentage of subsequent surveys not identifying the same deficiencies

Prior survey with both
condition- and
standard-level
deficiencies (n=271) 18.1 43.9 38.0

Prior survey with only
standard-level
deficiencies (n=1,802) n/a 28.6 71.4

n/a = not applicable

Source: GAO analysis of CMS OSCAR data.

ESRD surveyors in 6 of the 10 states in our study stated that they try to
reduce the occurrence of persistent noncompliance by taking a facility's
previous survey results into account when deciding which facilities to
survey. Following this policy, facilities doing poorly on one survey
should undergo a recertification survey more frequently. However, CMS's
current goals for ESRD surveys, because they focus solely on the frequency
of survey performance and not on the effectiveness of survey targeting,
create a disincentive for states to give greater attention to previously
noncompliant facilities. In particular, CMS's mandate to survey every
facility within a 3-year period tends to discourage survey agencies from
revisiting poorly performing facilities until all other facilities have
been inspected.

An analysis of survey activity from fiscal year 1998 through 2002
indicates that targeting of facilities based on their past survey results
occurred to only a limited extent in recent years. Only 5.9 percent of
facilities surveyed from fiscal year 1998 through 2001 with
condition-level deficiencies were resurveyed within a year, compared to
3.9 percent of facilities that had no condition-level deficiencies that
also were resurveyed within a year. The difference was somewhat greater
over a 2-year period, with 20.8 percent of facilities having
condition-level deficiencies in fiscal year 1998 through 2000 being
resurveyed compared to 12.6 percent of facilities that had no
condition-level deficiencies. Nonetheless, the large majority of
facilities with condition-level deficiencies were not resurveyed on an
accelerated 1-or 2-year schedule.

CMS Has Few Options to Sanction Noncompliant Facilities

State agencies are hampered in their ability to induce facilities to
comply fully and consistently with Medicare quality standards by the
paucity of sanctions available for cases of noncompliance. At present, the
only penalty that CMS can impose on ESRD facilities that do not comply
with these requirements is revoking their eligibility to participate in
the Medicare program. However, facilities typically are given a grace
period- usually 3 months-in which to correct any problems identified in a
survey. As long as these deficiencies have been addressed when surveyors
revisit the facility, the provider suffers no adverse consequences from
having failed to maintain compliance with Medicare quality standards.22
Consequently, very few ESRD facilities are terminated from Medicare, and
those that are can apply for readmission to the program. From fiscal years
1998 through 2002, only one dialysis facility was terminated from the
Medicare program and stayed out of business.23

Moreover, state survey agencies are often reluctant to press for the
termination of dialysis facilities because such closures would force
patients to find another provider and, in general, reduce patient access
to care. Many surveyors expressed a need to have additional sanctions
available to deal with poorly performing ESRD facilities. A number of such
alternatives already exist for nursing homes, including a denial of
payment sanction for new patients and civil monetary penalties. Denying
Medicare payments for new patients would curb the facility's major source
of revenue without eliminating, as a termination does, its ability to
serve existing patients. However, the lost revenue from potential new
patients, while the sanction is in effect, creates a concrete incentive
for the facility to resolve its quality problems quickly and to stay in
compliance thereafter. In addition, CMS requires states to refer for
immediate sanctions nursing homes found to have actually harmed one or
more residents or exposed them to potential serious injury on successive
surveys. In this situation, no grace period is granted to the facility.
Having

22If surveyors find that the facility is still out of compliance at the
first revisit, additional revisits are usually scheduled. Some facilities
get as many as four or five separate opportunities to demonstrate that
they have achieved compliance with Medicare's minimum quality
requirements.

23Adverse results on surveys could contribute to a provider's decision to
close a facility, even without a termination from Medicare. An examination
of OSCAR data for fiscal years 1998 though 2002 revealed six instances
where facilities closed voluntarily within 6 months of a survey that had
condition-level deficiencies. Five different facilities were recorded as
voluntary terminations, but remained open at the same addresses, sometimes
under new names and sometimes not.

multiple sanctions available means that surveyors can recommend the one
that best fits a given set of circumstances, taking into account the
likely impact on both the facility and the patients it serves.

In our June 2000 ESRD report, we noted that CMS had the authority to
expand the enforcement tools available for addressing quality problems
with ESRD facilities, but had not issued regulations and procedures to
implement alternative sanctions. Other sanctions, notably civil monetary
penalties, would require legislative changes by Congress. At that time, we
recommended that CMS act to expand available penalties where permitted
under its existing authority and that Congress consider authorizing civil
monetary penalties for dialysis facilities comparable to those already in
place for nursing homes. Since then, there have been no regulatory or
legislative actions to expand available enforcement tools for ESRD
facilities.

The publication of survey results could provide another incentive for
facilities to maintain compliance with Medicare quality standards. If ESRD
patients were able to readily compare the outcomes of surveys for
facilities in their area, they could choose to seek care from facilities
with more favorable inspection results. CMS has not taken any steps to
make survey results publicly available. By contrast, CMS routinely posts
survey results for nursing homes on an Internet Web site called Nursing
Home Compare. In 2001, when CMS created a comparable Web site covering
ESRD facilities, Dialysis Facility Compare, it chose not to make survey
results accessible.

The limitations inherent in state survey processes have been compounded by
inconsistent CMS oversight. On the one hand, CMS has substantially
increased funding for ESRD surveys in line with its expectation that
states survey a higher proportion of facilities each year. On the other
hand, survey agencies do not always receive the monitoring and technical
support that could enhance ESRD survey effectiveness. CMS regional offices
vary widely in the extent to which they examine states' ESRD survey
activities and provide related assistance. In addition, many state
agencies do not routinely have access to information from ESRD networks
that could assist them in selecting facilities to survey. Finally, the
limited number of CMS courses has made it difficult for many state
surveyors to obtain the training considered necessary to conduct ESRD
surveys.

CMS Has Increased Funding for State Surveys, but Monitoring and Technical
Support Are Uneven

Funding Has Increased to Support CMS's ESRD Survey Goals

In recent years, financial support for state survey activities overall has
grown substantially. According to the Director of CMS's Survey and
Certification Group, the increases responded to concerns that financial
support for survey activities was not keeping pace with the growth in
facilities and was putting Medicare beneficiaries at risk. From fiscal
year 1998 to 2002, total federal expenditures for state surveys increased
about 60 percent, with spending for long-term care (LTC) and non-LTC
facility survey activities growing 61 and 56 percent, respectively.24
Non-LTC facility survey activities are supported almost entirely by
federal funds, which must be allocated by states among home health
agencies, hospices, ambulatory surgical centers, rehabilitation
facilities, and other types of providers, as well as ESRD
facilities-within a set of guidelines established by CMS. ESRD survey
activities, therefore, must compete for funding with other non-LTC survey
activities, including statutorily-required surveys for home health
agencies that receive a higher priority. However, survey goals for ESRD
facilities are more ambitious than those for hospices, ambulatory surgical
centers, and many other non-LTC providers as CMS expects the agencies to
survey ESRD facilities more frequently.

Notwithstanding the competing survey priorities, the expansion in
financial support allowed state survey agencies to increase funding for
ESRD surveys to help meet higher survey goals. We estimated that federal
expenditures for ESRD survey activities nearly tripled from fiscal year
1998 to 2002, from $3.1 million to $8.2 million.25 Most of the increase
occurred between fiscal years 2000 and 2001, when the ESRD survey goal
almost doubled from 17 to 33 percent of a state's facilities each year.
(See table 4.) Increased spending for ESRD survey activities was evident
across nearly all states. From fiscal year 1998 to 2002, 42 states had an
increase in spending for ESRD survey activities, and the median state
experienced a 144 percent increase.

24CMS allocates most funding for state survey activities by LTC and
non-LTC categories. LTC funding covers surveys of nursing homes and
ICF/MRs. Non-LTC funding supports surveys of dialysis facilities, home
health agencies, accredited and nonaccredited hospitals, hospices,
ambulatory surgical centers, outpatient physical therapy providers, rural
health clinics, comprehensive outpatient rehabilitation facilities,
portable x-ray suppliers, psychiatric residential treatment facilities,
and psychiatric hospitals.

25These estimates are based on workload and expenditure reports provided
annually to CMS by state survey agencies, which combine all non-LTC survey
activities. Several state governments also fund provider survey and
certification activities for non-LTC providers. In fiscal year 2001, state
support accounted for approximately 6 percent of total spending on non-LTC
activities.

Table 4: Federal Support for Provider Surveys, Fiscal Years 1998 to 2001

Dollars in millions

Non-long-term care provider surveys

Total

Long-term care provider surveys

Non-ESRD surveys (estimated)

ESRD surveys (estimated)

1998 $ 253.2 $ 209.2a $ 41.0a $ 3.1a
1999 265.1 217.2b 44.5b,c 3.3b,c
2000 312.1 260.3 47.3d 4.6d
2001 350.6 288.9e 53.7e 8.1e
2002 405.2 336.6 60.5

Source: CMS aggregate budget data for Medicare and Medicaid survey
activities.

Note: GAO estimates are based on the ESRD share of non-LTC survey hours
reported to CMS. The three budgetary subcomponents do not sum to totals
because of rounding.

aExcludes Nebraska.

bExcludes Tennessee.

cExcludes Washington.

dExcludes Arkansas.

eExcludes Vermont and Virginia.

In most states, the increase in ESRD spending outpaced the growth in
spending for all non-LTC survey activities. As a result, the ESRD share of
non-LTC expenditures also increased, from about 7 percent of non-LTC
survey expenditures in fiscal year 1998 to about 12 percent in fiscal year
2002. For fiscal year 2002, we estimated that the ESRD share of non-LTC
survey expenditures across states ranged from about 0 to 35 percent. For
the states with the largest number of dialysis facilities, the ESRD share
ranged from 6 percent in Virginia to 25 percent in Georgia.

Regional Office Monitoring Regional offices' review of agency surveys,
referred to as federal and Assistance to State monitoring surveys, are
conducted by CMS to monitor state agency Agencies Are Highly performance
in interpreting and applying federal standards as well as to Inconsistent
identify training or technical assistance needs of surveyors. Although CMS

is required to conduct monitoring surveys that assess the adequacy of the
state's survey for nursing homes, no similar legislative requirements
apply

to ESRD facilities.26 As such, CMS has used monitoring surveys for
dialysis facilities that are observational in nature-regional office staff
accompany state surveyors on inspections of dialysis facilities, observe
them as the surveyors identify and document facility deficiencies, and
provide feedback on the surveyors' performance. CMS has not specified the
number of ESRD monitoring surveys that regional offices should conduct.
Perhaps as a consequence, representatives for six regional offices that we
contacted-responsible for 29 states-told us they have conducted very few
such surveys over the last 2 fiscal years. In fiscal year 2001, the number
of monitoring surveys each regional office performed ranged from 3 to 11;
in fiscal year 2002, they ranged from 2 to 6. None of the regional offices
in either year conducted a monitoring survey for every state in its
jurisdiction.

Even for the few monitoring surveys conducted, most CMS regional offices
in our study provided little feedback to the states. At 3 of the 10 state
survey agencies we contacted, representatives reported receiving only one
monitoring survey in 5 years and were provided no feedback. Other survey
agency representatives stated that regional offices provided verbal
feedback on their monitoring surveys. In contrast, two CMS regional
offices also provided written feedback that included evaluations of
surveyors' decisions regarding specific conditions and standards.

The regional offices in our study also have not taken full advantage of
available data to monitor state agencies' survey performance for ESRD
activities. CMS has instructed regional offices to use data from its OSCAR
system as an integral tool to assess and compare state agency performance,
particularly differences in the time required to conduct surveys and the
types of deficiencies cited. According to CMS, such analyses can provide
the information necessary to help state agencies improve their efficiency
in conducting ESRD surveys and achieve consistency in their quality. For
example, because OSCAR contains data on the number of hours spent on each
ESRD survey, regional offices could use a benchmark to compare and assess
survey times across their state agencies. CMS has indicated that similar
analyses could be performed for the types of deficiencies cited by
surveyors to determine whether there

26For each state, CMS is required to perform validation surveys-on-site
inspections of facilities, separate from those conducted by the state
agency-for at least 5 percent of the nursing home surveys conducted
annually, but no fewer than five homes in each state. See 42 U.S.C. S:
1395i-3 (g) (3) (B) (2000).

were any differences in state agencies' application of quality
standards.27 Despite such potential uses of data to monitor state agency
performance, most of the regional offices analyzed their available data on
a more limited basis. They checked on past survey results for certain ESRD
facilities and relied extensively on quarterly workload reports from each
state agency to determine the number of recertification surveys conducted.

In addition to monitoring and tracking ESRD survey activities, CMS
requires regional offices to assist state agencies in fulfilling their
survey responsibilities. Such assistance includes alerting the agencies to
CMS policies and goals, coordinating communications with the CMS central
office, helping surveyors obtain ESRD training, and consulting on a
regular basis on program activities and achievement of survey goals. The
performance of regional offices in our study varied from little contact
with their state agencies to extensive collaboration. One CMS regional
office had almost no contact with its state survey agencies or network and
was not sure of the state agencies' performance in meeting ESRD survey
goals. A survey agency representative in that region stated that contact
with the regional office consisted primarily of a few calls the agency
made to obtain clarification on a policy or procedure. In contrast, most
of the regional offices included in our study, at a minimum, contacted
state survey agencies to discuss CMS policies and goals, provided
technical information or training on ESRD issues, and offered assistance
in conducting select surveys.

Among the most active regional offices in providing support on ESRD
surveys was Region 9.28 Its efforts to improve state agency survey
performance included a range of activities:

o The office collaborated with state agencies and networks to provide
ESRD training to state surveyors in addition to that provided by the CMS
central office.

o Through conference calls, the office contacted its state agencies
monthly (including their district offices) to discuss current ESRD survey
issues, relevant federal bulletins or alerts, instructions for more
consistent coding of deficiencies, updates on training needs and slots
available, and surveyor

27Regional offices have used OSCAR data to prepare tracking reports on
areas related to state and regional office performance for nursing home
surveys, including facility terminations, number of surveys without
deficiencies, and analyses of most-frequently cited deficiencies across
states.

28Region 9 includes state survey agencies for Arizona, California, Hawaii,
and Nevada.

decisions related to inspection findings. The conference calls provided a
mechanism for surveyors to pose questions directly to CMS officials and
often receive an immediate response.

o The office conducted quarterly conferences that included
representatives from the networks and state survey agencies to provide
updates on quality improvement programs underway by the networks, general
issues related to ESRD, and issues specific to certain facilities.

o The office joined state agencies and networks in a campaign to educate
facility managers about ESRD regulations and the survey process.

Disparities in regional office performance-not unlike the disparities in
state survey agency performance-may reflect their ability to cope with
CMS's survey priorities. Officials representing several regional offices
noted that CMS's focus has been on nursing homes and other types of
facilities that are a higher survey priority than ESRD facilities. Some of
these officials indicated that, as a consequence, needed attention in
monitoring state agencies and providing technical assistance for ESRD
survey activities has lagged.

Networks Do Not Routinely Share Facility Data with State Agencies

State survey agencies are not routinely receiving information from ESRD
networks-organizations authorized by statute to collect information on
patient complaints, quality improvement projects, and clinical
performance. The networks operate under contracts with CMS which, in
fiscal year 2002, totaled $24.7 million, approximately three times the
amount of federal funds we estimate were spent on state survey and
certification activities for dialysis facilities.29 Networks use the
information they collect to perform a wide range of quality improvement
activities and to identify and address any quality issues that may arise
with individual facilities. Under the terms of their CMS contract, they
are to cooperate with state survey agencies by providing them
facility-specific information upon request. However, our June 2000 study
found that most CMS regional offices had restricted networks from sharing
facility-specific information, contending that federal confidentiality
regulations prohibited such exchanges. In response, we recommended that
CMS establish procedures

29Network responsibilities are established by the Social Security Act,
which also authorizes the Secretary to prescribe other network duties and
functions. SeeS: 1320c-9(b)(1) and S: 1395rr(c)(2). Current network
responsibilities are set forth in contract: ESRD Network Organizations,
Statement of Work, FY 1999-2003, Section C.4.F, Cooperative Activities
With State Survey Agencies and Quality Improvement Organizations, CMS. The
networks are funded through a fifty-cent charge on each Medicare dialysis
treatment.

to facilitate routine cooperation and information sharing between networks
and state agencies. The HHS Inspector General made similar recommendations
in June 2002.30 However, most of the states in our current review reported
that they have seen little evidence of increased information sharing by
ESRD networks.

Most of the state agencies included in our study did not receive
facility-specific information from networks on a regular basis. State
agency officials indicated that the networks typically provided summary
data for facilities, and that access to facility-specific information
occurred on a case-by-case basis. Much of the information that was shared
by networks came in response to inquiries from state agencies regarding
specific providers. In addition, networks rarely identified facilities as
candidates for inspection. For example, one state agency official noted
that the area ESRD network rarely shared information on complaints and
made only one recommendation over the last 5 years that identified a
facility for inspection.

Several state agency officials attributed the limited disclosure of
facility-specific information to confusion in the ESRD community about
requirements pertaining to safeguarding this information. The Social
Security Act prohibits the disclosure of facility-specific information to
any person subject to several exceptions, for example, where federal
regulation authorizes the disclosure in order to protect the rights and
interests of patients.31 Although their contracts with CMS indicated that
the agency wanted them to share facility-specific information with state
survey agencies, the networks are hesitant to follow this directive
because the agency regulations do not identify such disclosure as a
specific exemption from the general statutory prohibition. Reportedly,
network officials are concerned that the release of such information could
undermine their quality improvement efforts and collaborative
relationships with facilities.32 CMS acknowledged that confusion exists in
this area and convened a workshop to promote more understanding and

30Department of Health and Human Services, Office of Inspector General,
External Quality Review of Dialysis Facilities: A Call for Greater
Accountability, OEI-01-00050 (Washington, D.C.: June 2000).

31See42 U.S.C. S:S: 1320c-9(a) and (b).

32CMS policy stipulates that state agencies may not release confidential
information that they receive from ESRD networks to third parties, even
under subpoena.

cooperation between the networks and the state agencies.33 However, CMS
has not required networks to routinely share facility-specific
information.

The potential benefits that can be achieved from increased sharing of
network information are well illustrated by the recent experience of the
California state survey agency. The state agency routinely receives
facility-specific information from its two corresponding networks
verbally-no facility-specific data are sent to the state agency in written
form. Regardless of the method, the networks and the state agency agreed
that they need to be able to share such information, considering its
potential benefits in improving facilities' quality of care, and have
conveyed this to the ESRD facilities' managers. Consequently, the networks
regularly contact the state agency to share different types of quality of
care information on individual facilities, including complaints the
network received and the results of related investigations. The networks
now routinely make suggestions regarding potential facilities for the
state agency's attention. This relationship improved markedly after years
of little communication between the state agency and the networks, largely
as the result of increased trust derived from working together on a series
of joint projects.34

States Report Insufficient ESRD Training Opportunities

According to state officials, scarcity of ESRD training opportunities has
impeded state agencies' efforts to improve surveyor performance. Because
most surveyors do not have prior training or experience in dialysis, state
survey agencies have for years relied on the courses that CMS has
organized to train ESRD surveyors in the technical aspects of dialysis and
the application of ESRD quality standards. The need for specialized
training is consistent with the highly technical nature of ESRD surveys
relative to surveys of other provider types. CMS offers basic ESRD
training for surveyors who are not experienced with ESRD surveys and
advanced training for others. Officials at the state agencies in our study
generally

33To encourage data sharing, CMS has begun work on a draft Memorandum of
Understanding that the state agencies and ESRD networks could adopt.

34In 2000, the state agency and the network participated in a special
project intended to increase the number and quality of ESRD surveys. The
networks, along with the state agency and the CMS regional office, jointly
provided ESRD training to surveyors who had limited experienc with ESRD
surveys. All three then worked together to help facilities correct
deficiencies and have since collaborated on educating facility managers
about ESRD standards and the survey process.

commended these courses, noting that they provided surveyors with the
knowledge and skills needed to conduct ESRD surveys effectively.

Three of the state agencies we reviewed require that surveyors complete
CMS's basic ESRD training before they are allowed to perform surveys
unassisted. State agency officials emphasized that they try to get
surveyors trained as quickly as possible after they have been assigned
ESRD survey responsibilities. This not only permits the surveyors to gain
expertise in conducting ESRD surveys at the appropriate time, but it also
allows them to begin conducting surveys unassisted in a timely fashion.
Delays in getting surveyors scheduled for basic training delays their
readiness to conduct surveys unassisted, which in turn has an impact on a
state agency's performance in the number of surveys it conducts during the
year.

For most of the state agencies in our study, the limited number of CMS
training classes offered has been problematic. In particular, the
infrequency of classes at the introductory level for ESRD training has had
the greatest impact on state agency operations. From fiscal year 1999 to
2002, CMS offered only one course each year for basic training, always
given at the same time of year, and since fiscal year 2000, always in
Denver. In light of this schedule, state agencies were particularly
concerned about the delay in training surveyors who were new to ESRD. At
times, state agencies sent these surveyors to take advanced courses when
openings in the basic course were unavailable. However, these courses
dealt largely with selected topics and did not explain the core technical
and regulatory concepts covered in the basic course. As a result,
surveyors who had previously taken basic training and had some experience
in conducting ESRD surveys found the advanced courses most informative and
useful. Officials of several state agencies also indicated that CMS could
help accommodate surveyors by offering basic ESRD training at multiple
sites, taking into consideration the location of class enrollees. Some
officials added that this would provide the additional benefit of helping
their agencies save funds used for travel.

CMS has highlighted the value to surveyors of attending its basic ESRD
training course by instituting a new policy that requires all newly
appointed ESRD surveyors to complete it. Effective fiscal year 2003, all
newly hired ESRD surveyors, or surveyors who have not previously performed
ESRD surveys, must complete the course before they can serve in a capacity
other than a trainee. However, CMS has chosen not to fill this gap for
surveyors who took advanced courses as a substitute for the basic course
in years past. For surveyors who performed ESRD surveys prior to

Conclusions

fiscal year 2003, other CMS ESRD training courses are considered
equivalent. Experienced ESRD surveyors who have not received any ESRD
training from CMS have until fiscal year 2004 to complete either ESRD
basic or advanced training.

CMS fielded a questionnaire to state agencies to determine current
training needs in light of the new training requirement. Although the
results of this survey are still being reviewed and analyzed by CMS,
preliminary tabulations indicate that at least 21 percent of experienced
ESRD surveyors met the training requirement through one of the presumed
equivalent courses and had never taken the CMS basic course. In at least
six states fewer than half the surveyors had taken the basic ESRD
training. The extent to which experience in conducting ESRD surveys
compensates for a lack of formal training is an open question. Until that
process is complete, the scarcity of training opportunities in the past
could continue to constrain the effectiveness of many ESRD surveyors.

As a result of critical weaknesses in the system established to monitor
and enforce compliance with Medicare's quality standards for ESRD
facilities, full and consistent compliance with these standards has become
more the exception than the rule. Despite increased surveying goals
recently set by CMS, many facilities continue to escape the attention of
state surveyors for long periods of time. This is especially problematic
for facilities that have performed poorly in the past and are therefore
relatively more likely to reveal deficiencies when surveyed again. In
addition, there are few if any negative consequences for facilities if
they are surveyed and found out of compliance with Medicare's quality
standards. Currently, facilities can escape negative publicity from having
multiple deficiencies, despite the fact that the statement of deficiencies
prepared by state surveyors is a public document.

The wide variation across states in the number of condition-level
deficiencies found indicates in part that some surveyors are more
proficient than others in detecting quality problems. ESRD survey
expertise can be enhanced through training and experience. Promoting
surveyor specialization should lead to more thorough ESRD inspections and
more accurate documentation of deficiencies. Similarly, were CMS to offer
more basic level ESRD courses, at different locations and times, surveyors
newly assigned to ESRD facilities could more quickly obtain the training
they need to conduct effective inspections. In addition, a comparable
expansion in advanced course offerings would enable a larger

proportion of experienced surveyors to catch up with technical
developments in dialysis treatments.

State survey agencies could better target their survey activities if they
had access to information from ESRD networks on the extent of serious
quality problems at individual facilities. However, CMS regulations that
require networks to safeguard the confidentiality of data that they obtain
from dialysis facilities has generated confusion among the networks as to
what facility-specific information they legitimately can and should share
with state survey agencies. CMS could remove this long-standing impediment
by revising those regulations to clearly make such data sharing with state
agencies mandatory.

Moreover, the magnitude of variation across states in the level of survey
activity and survey results underlines the need for more intensive
monitoring of, and support to, the individual state agencies. However, CMS
has not addressed the enormous variation among its own regional offices in
the extent to which they undertake these activities. The highly
inconsistent performance in the number of ESRD surveys conducted by state
agencies and surveyors' detection of deficiencies may reflect uneven
monitoring and support provided to them by CMS regional offices-some of
which devoted considerable attention to ESRD survey activities, and
others, virtually none.

Ultimately, no quality assurance system can be effective unless providers
face real consequences when they are cited repeatedly for deficiencies.
Because they are routinely given multiple opportunities to demonstrate
that they have corrected any problems found, ESRD facilities have no
strong incentive to adhere to those standards until a survey takes place.
Facilities are likely to continue cycling in and out of compliance until
state agencies have a broader range of enforcement tools, especially ones
that take effect even if deficiencies are subsequently corrected. CMS
could implement some additional sanctions by regulation. However, as we
noted in our June 2000 report, CMS did not have the authority to expand to
ESRD facilities the range of alternative sanctions available for use
against noncompliant nursing homes. We therefore suggested at that time
that Congress consider authorizing CMS to impose civil monetary penalties
on dialysis facilities. Our current work supports consideration of this
suggestion.

Moreover, the effectiveness of alternative sanctions would be greatly
strengthened if they could also be imposed promptly, without allowing
facilities a grace period to correct identified deficiencies. Such
immediate

sanctions could be applied when facilities are found to have
condition-level deficiencies in successive surveys. For instance,
immediate denial of payments for new patients could create a strong
incentive to maintain compliance because the facility loses income from
Medicare, which usually represents a substantial part of operating
revenues.

To encourage ESRD facilities to sustain their compliance with Medicare
quality standards, Congress should consider authorizing CMS to immediately
impose a sanction when a dialysis facility has condition-level
deficiencies in successive surveys without providing the facility a grace
period before the sanction takes effect. The immediate sanction options
available to CMS should include denial of Medicare payments for new
patients and civil monetary penalties.

Matter for

Congressional
Consideration

Recommendations for We recommend that:

Executive Action To create incentives for facilities to maintain
compliance with Medicare quality standards, the Administrator of CMS
should

o establish a goal for state agencies to reduce the time between surveys
for facilities with condition-level deficiencies and

o publish facilities' survey results on its Dialysis Facility Compare Web
site.

To help surveyors identify and systematically document deficiencies, the
Administrator of CMS should

o strongly encourage states to assign ESRD inspections to a designated
subset of surveyors who specialize in conducting ESRD surveys and

o make ESRD training courses more available to state surveyors, which
may include increasing the number of classes and slots available as well
as varying class location.

To enhance the support and monitoring of state survey agencies, the
Administrator of CMS should

o amend its regulations to require that networks share facility-specific
data with state agencies on a routine basis and

o ensure that regional offices both adequately monitor state performance
and provide state agencies ongoing assistance on policy and technical
issues through regularly scheduled contacts with state surveyors.

Agency Comments
and Our Evaluation

In its written comments, CMS did not indicate an intention to implement
five of our six recommendations. Nevertheless, it affirmed its commitment
to ensuring adequate oversight of dialysis facilities and state survey
agencies, and described a number of measures that it has initiated to
strengthen this process. (CMS's comments are reprinted in app. IV.)
However, two of these initiatives-a proposed survey of ESRD beneficiaries
and the automation of data reporting by facilities to CMS- will only
indirectly affect the survey and certification program that was the focus
of our report. In our report, we identified several key limitations in the
structure and implementation of this program that constrain its
effectiveness in enforcing Medicare's quality standards for ESRD
facilities. In addition to comments on each of our recommendations, CMS
also provided technical comments that we incorporated where appropriate.

With respect to our matter for congressional consideration, CMS affirmed
its commitment to take action against ESRD facilities with serious quality
problems. It also acknowledged that the agency needed to create strong
incentives for facilities to provide quality care. The agency proposed to
address this issue by initiating an evaluation of the effectiveness of
sanctions on improving nursing home care. Although such an evaluation may
produce useful information about nursing homes, it will have limited
relevance for the quality of care provided to ESRD patients. We continue
to believe that Federal oversight of dialysis facilities could be improved
by strengthening the enforcement process. Therefore, Congress should
consider authorizing CMS to impose immediate sanctions on dialysis
facilities cited with serious deficiencies in consecutive surveys.

CMS's response to the first of our recommendations for executive action-
that it set a goal for more frequent surveys of facilities with a history
of condition-level deficiencies-acknowledged the value of targeting
surveys on poorly performing providers. Though it expressed a strong
commitment to increased oversight of such facilities, CMS did not indicate
a willingness to set this additional goal. Instead, CMS relies on the
states to use the flexibility that it has built into its budget call
letter to target their surveys on ESRD providers most likely to have
quality problems. However, we found that the budget call letter placed
ESRD facilities in a lower priority category, behind both nursing homes
and home health agencies. Without a change in the priorities that CMS has
communicated to the state agencies, it is unrealistic to expect most
states to go beyond the goals currently set by CMS for ESRD survey
activity.

In its comment, CMS also highlighted its efforts to develop tools to help
state agencies identify facilities that are most likely to exhibit quality

problems. These include reports on individual facilities-produced from
claims data and other administrative data files by CMS contractors-that
describe their practice patterns and outcomes. CMS also stated that it
distributes to the states an Outcomes List that ranks facilities for
surveying priority based on their performance on dialysis adequacy, anemia
management, and adjusted mortality rates. However, CMS's surveying goals
for the states, as they are currently structured, do not focus on
targeting of any sort. Our analysis of state survey activity found scant
evidence that state agencies were conducting more frequent surveys of even
the most obvious candidates-facilities that had condition-level
deficiencies in their most recent prior survey. Our evidence and CMS's
response indicates a need for CMS to go beyond its current efforts to
developing inspection goals on poorly performing facilities.

CMS did not directly respond to our second recommendation, that CMS
publish survey results on its Dialysis Facility Compare Web site. Instead,
the agency described various studies it has underway to develop better
information for consumers, including efforts to make survey results more
uniform across the country. While greater uniformity in survey results is
a laudable objective, we would note that the results of surveys currently
conducted are the basis for the agency's decisions to either recertify or
(potentially) terminate ESRD facilities as Medicare providers. Therefore,
the information we have recommended that CMS share with the public does
not represent an abstract quality indicator of unknown validity. Rather,
it conveys the actual status of the facility in terms of fulfilling its
basic obligation to meet Medicare's conditions for coverage. In our
opinion, these nominally public, but heretofore undisseminated, survey
outcomes would convey useful information to interested ESRD patients
trying to decide among alternative facilities.

Our third recommendation was that CMS encourage state agencies to identify
a subset of surveyors who would specialize in conducting ESRD facility
inspections. In its comment CMS did not address our recommendation but
responded that, in general, it encouraged states to have specialized
surveyors when possible. However, the agency did not describe what
specifically it had done to promote this practice. CMS did highlight other
initiatives it has taken to enhance surveyor skills and improve the survey
process more generally. These include its development of a new software
system to help guide surveyors as they conduct surveys, the reports on
practice patterns and outcomes of individual facilities, and increases in
the surveyor training that CMS provides. CMS concluded that these steps
were the most appropriate use of limited resources. We would note,
however, that to the extent that states do not concentrate their ESRD

surveys on a subset of specialist surveyors, more surveyors will need to
receive CMS training in conducting ESRD surveys. That represents a less
efficient use of CMS training resources. We continue to believe that
surveyor specialization contributes to more thorough and effective
inspections, in addition to whatever benefits accrue from other
improvements such as expanded training and customized software.

Our fourth recommendation was that CMS expand the number and slots
available in training courses for ESRD inspections, as well as vary their
locations. CMS responded that it has arranged to increase its offerings to
a minimum of two basic ESRD training classes annually, with one course
conducted in Denver and one in Minneapolis. According to CMS, more
advanced ESRD training may also be increased, depending on demand. This
expansion should lessen considerably the difficulty that state survey
agencies have experienced obtaining the necessary training for their ESRD
surveyors on a timely basis.

In our fifth recommendation, we urged CMS to amend its regulations to
require that ESRD networks share facility-specific data with state
agencies on a routine basis. CMS responded that networks are currently
required to share data with CMS, which can then provide appropriate
information, such as the previously mentioned Outcomes List, to state
agencies. CMS also stated that information that networks obtain through
their quality improvement efforts has limited utility for quality
assurance because it is not standardized (that is, the specific
information collected will vary across networks and projects). On the
contrary, we found that the networks' quality improvement projects collect
new information directly from dialysis facilities which helps identify
those facilities that perform poorly on one or more quality dimensions. As
the experience of California has shown, such data can provide valuable
guidance to state surveyors in their selection of facilities to inspect,
regardless of whether identical information is collected by every network
across the country.

Our last recommendation stated that CMS should ensure that its regional
offices provide adequate oversight of, and assistance to, state agency
monitoring of ESRD facilities. As with several previous recommendations,
the agency reaffirmed its commitment to the overall goal, but did not
address the weaknesses that we found in its implementation. CMS's comment
describes the resources available to the regional offices, including
assigned ESRD specialists, regional data reports, and monthly conference
calls with state agency officials. However, CMS did not address the large
variation across regions in the extent to which they use these tools, and
refers to no specific measures intended to stimulate greater

effort in regions that have been less active to date. CMS stated that it
is working hard to clarify its expectations for both state agencies and
its own regional offices, but in its comment provides no explanation or
examples of what this might entail.

As agreed with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution of it until 30 days
from
its date. At that time, we will send copies of this report to the
Administrator of CMS and to other interested parties. In addition, this
report will be available at no charge on GAO's Web site at
http://www.gao.gov. We will also make copies available to others upon
request.

If you or your staff have any questions about this report, please call me
at
(312) 220-7600. An additional GAO contact and other staff members who
prepared this report are listed in appendix V.

Sincerely yours,

Leslie G. Aronovitz
Director, Health Care-Program

Administration and Integrity Issues

Appendix I: Scope and Methodology

Quality of Care

To analyze variation in the clinical performance of individual facilities,
we downloaded information available from CMS's Web site, Dialysis Facility
Compare (DFC)-http://www.medicare.gov/Dialysis/Home.asp.1 DFC provides
information on two clinical performance measures: the proportion of
patients with adequate hemodialysis-defined as a Urea Reduction Ratio of
at least 65-and the proportion of patients with adequate anemia
control-defined as a hematocrit of 33 or better. DFC has data on the
latter measure for patients taking the drug erythropoietin (EPO)-the
therapy generally used to treat anemia among ESRD patients. The most
currently available data for both measures came from information provided
on Medicare claims submitted for treatment furnished in 2000. DFC reports
the proportion of patients at each ESRD facility who achieved the
designated threshold for these two measures.

To provide a more concrete sense of the types of quality problems
encountered by state surveyors, we selected five survey reports, known
formally as a "statement of deficiencies" (Form 2567), that described in
detail the deficiencies cited in inspections of individual facilities in
five states. We abstracted from each survey report the justification
written by the surveyor for one deficiency citation. The episodes we chose
involved deficiency codes that are widely cited among survey reports
nationwide. In the data we assembled from CMS's Online Survey
Certification and Reporting (OSCAR) system, at least one of these six
specific deficiency codes-111, 112, 118, 240, 264, and 423-was cited in 35
percent of all recertification surveys conducted in fiscal years 1998
through 2002.

To more fully appreciate the clinical consequences of these deficiencies
for patients, we shared our abstracted citations with three ESRD
surveyors, each with at least 5 years of ESRD survey experience, whom we
had previously interviewed in conjunction with our site visits to three
different states. All were registered nurses. The three surveyors
commented on each of the six vignettes that we sent them by describing the
potential impact of these situations on patient health and well-being.
Their analyses encompassed expected symptoms, such as fatigue, swelling,
and shortness of breath, medical conditions that could result, such as
heart failure and ruptured red blood cells, and related outcomes, such as
shortened life expectancy.

1Other clinical performance measures have only been reported from samples
of patients, providing data on national and regional trends but without
the ability to compare results across individual dialysis facilities.

Appendix I: Scope and Methodology

Survey Frequency and Results

To analyze the frequency and results of surveys conducted in the 50 states
plus the District of Columbia, we obtained all the data stored on CMS's
OSCAR system relating to standard surveys of ESRD facilities. Standard
surveys include initial surveys-conducted when a facility first applies
for Medicare certification-and recertification surveys-conducted at
intervals subsequent to the initial survey for that facility.2 The OSCAR
database is continuously updated and retains data for the four most recent
surveys for each facility. Our analysis was not adversely affected by the
potential loss of data if a given facility had more than four standard
surveys conducted, because less than 1 percent of ESRD facilities had as
many as four surveys from fiscal year 1998 through 2002, the period of our
review.

When state survey agencies complete their work on these surveys, CMS
requires them to record in OSCAR information about the inspection
including the dates that the surveys took place and the specific
deficiency codes for each standard-level and condition-level deficiency
cited. OSCAR also contains Provider of Service file information on ESRD
facilities, including their name, address, chain ownership, date of
Medicare enrollment, and the date of, and reason for, termination (if
any).

The data used in our OSCAR analyses was downloaded on April 2, 2003,
providing a 6-month period following the end of fiscal year 2002 for state
agencies to complete the process of data entry. To assess the completeness
of the data, we compared the number of surveys we found in OSCAR for
fiscal years 1998 through 2002 with the number of surveys that state
agencies indicated that they completed in annual workload reports
submitted to CMS. Although complete workload data were not always
available, where they were, the numbers of ESRD surveys reported for most
states matched the number recorded in OSCAR either exactly or nearly (plus
or minus 3) in each of the 5 fiscal years.

In analyzing the proportion of ESRD facilities resurveyed in fiscal years
1998 through 2002, we determined the facilities that were available for
recertification in each year. We excluded those facilities that were
subject to an initial survey, and any that had either dropped out of
Medicare prior to that year or that did not begin participating in the
program until later.

2Similar information is collected on complaint surveys but stored in
separate data files.

Appendix I: Scope and Methodology

Surveyor Specialization

To assess the effect of surveyor specialization, we analyzed the
relationship of survey results statewide with the degree of surveyor
specialization in that state. We defined specialization as assigning ESRD
facility inspections to a subset of surveyors who spend much of their time
focused on ESRD quality issues. We knew from our state site visits and
interviews that some states promoted specialization while other states
distributed ESRD assignments roughly equally among surveyors who spent
most of their time inspecting nursing homes and home health agencies. From
9 of the 10 states that we examined most closely, we obtained data on the
number of ESRD and non-ESRD surveys conducted by each surveyor during
fiscal year 2002. (We were not able to obtain this information from
Mississippi.) From those data we calculated the proportion of total
surveys that were of ESRD facilities, first for each individual surveyor,
and then for the state as a whole. The statewide ratio combined the
individual surveyor ratios, with each individual's ratio weighted by the
proportion of ESRD surveys in fiscal year 2002 accounted for by that
individual. The result was a state specialization score that had a
possible range from almost 0 to 1. (A state would get a score of 1.0 if
all of its ESRD surveys were done by surveyors who never inspected any
other provider types.) This approach was designed to gauge the relative
likelihood that any given ESRD survey in the state would be conducted by a
surveyor whose survey activities focused on ESRD facilities.

We assessed the strength of the relationship of surveyor specialization to
survey results by comparing the aggregate results of states with low
specialization scores with states that had relatively high scores.
Specifically, we compared the proportion of surveys with condition-level
deficiencies and the number of standard-level deficiencies cited in
surveys. We applied chi-square tests to determine if observed differences
between the two groups were likely to have occurred by chance, using the
conventional 95 percent confidence interval. We were not able to link the
results of individual surveys to the experience level of the surveyors who
conducted them. Therefore, our analysis compared aggregate survey outcomes
across two groups of states, distinguished by their overall level of
surveyor specialization.

Surveyor Training To assess the extent of state surveyor training to
perform ESRD facility inspections, we drew on the results of a survey
conducted by CMS of the state survey agencies. CMS solicited data on the
titles and dates of all CMS-sponsored training on ESRD completed by each
of the states' individual surveyors who had performed ESRD inspections
prior to fiscal year 2003. It initially collected these data in January
and February of 2003

Appendix I: Scope and Methodology

and continued obtaining updated and corrected information through May
2003. We analyzed the most recent data supplied to us by CMS at that time.

Federal Funding for ESRD Surveys

To assess federal funding for ESRD and other survey activities, we
reviewed quarterly and annual expenditure reports submitted to CMS by
state survey agencies for fiscal years 1998 to 2002. These reports specify
the funds spent by state agencies for both long-term care (LTC) and
non-LTC survey activities under the Medicare and Medicaid programs.
However, because the reports aggregate expenditures for all non-LTC survey
activities, we had to estimate the expenditures related to ESRD surveys.
We developed our estimates based on additional CMS data that indicated the
number of hours each state agency reported was spent on activities related
to ESRD surveys, as well as activities related to non-LTC surveys overall.
We then calculated the ESRD-related share of non-LTC survey hours and
applied that percentage to the total non-LTC survey expenditures each
state agency indicated on its annual expenditure report.

Appendix II: Medicare Conditions for Coverage for Dialysis Facilities

Condition for coverage Description

Compliance with federal, state, and local The facility and personnel
employed by the facility must be licensed as required by

laws and regulations federal, state, or local laws. This includes
compliance with all public safety laws and requirements.

Governing body and management The facility must be under the control of
an identifiable body that adopts and enforces rules and regulations,
including operational rules and patient care policies to safeguard the
health and safety of individuals.

Patient long-term-care program and A professional, multidisciplinary
health care team and the patient must develop a written

patient care plan long-term-care plan to ensure each patient receives the
appropriate type of dialysis and care. Patient care plans, which have
shorter time lines, must be personalized for each patient to address their
specific medical, psychological, social, and functional needs. Both plans
are to be regularly reviewed and updated to respond to changing patient
needs.

Patients' rights and responsibilities Dialysis facilities must have
written policies describing the rights of the patients in order to ensure
patients are fully informed about the services available, their medical
condition, whether the facility reuses dialysis supplies, and whether the
patient is a candidate for transplantation and home dialysis.

Medical records Patient medical records must be maintained to document
patient assessments, diagnosis, and treatment, and medical and nursing
histories.

Physical environment Dialysis services are to be provided in a setting
that is functional, sanitary, safe, and comfortable for patients, staff,
and the public.

Reuse of hemodialyzers and other dialysis supplies

Facilities that reuse hemodialyzers and other dialysis supplies must
follow established protocols and standards to ensure patient and staff
safety.

Affiliation agreement or arrangement Agreements between dialysis
facilities and inpatient dialysis centers must be in writing to ensure
inpatient care and other hospital services are promptly available to
dialysis patients.

Director of renal dialysis facility Dialysis treatments must be under the
general supervision of a qualified director, who is responsible for
planning, organizing, conducting, and directing professional services.

Staff of a renal dialysis facility or center Properly trained and
qualified personnel must be present in adequate numbers to meet the needs
of patients, including needs arising in emergencies.

Minimal service requirements Dialysis facilities must provide dialysis
services as well as laboratory, social, and dietetic services needed to
address ESRD patient needs.

Source: 42 C.F.R. Part 405 Subpart U (2002).

Appendix III: State Agencies' Progress toward Meeting CMS Survey Goals

The table below shows the percentage of facilities surveyed, by state, in
fiscal years 1998 to 2002. It indicates how the individual states
responded to the increases in the goal for annual ESRD recertification
rates set by CMS, from 10 to 17 percent per year in fiscal year 2000 and
then to 33 percent each year starting in fiscal year 2001.

Table 5: ESRD Facilities Recertified Annually by State, Fiscal Years 1998 to
2002

Percentage

1998 1999 2000 2001 2002

CMS goal 10 10 17 33

State recertification survey rates

Alabama 12 15 17 9
Alaska 50 0 100 100
Arizona 8 13 11 28
Arkansas 11 50 16 32
Californiaa 2 4 12 12
Colorado 7 6 24 6
Connecticut 17 28 27 39
Delaware 0 9 0 15
District of Columbia 8 4 28 22
Floridaa 10 14 22 37
Georgia a 12 12 17 37
Hawaii 21 13 6 25
Idaho 17 13 14 14

Illinois 29 21 22 41 32 Indiana 5 19 16 33 31 Iowa 911 17 2627 Kansas 14
24 21 44 33 Kentucky 63 54 70 83 89

Louisiana 10 18 17 32 31 Maine 3838 8 3325 Maryland 17 16 5 26 28
Massachusetts 8 13 14 37 31 Michigana 31 28 11 18 10 Minnesota 21 4 5 27
33 Mississippi 8 9 36 69 9 Missouri 5 7 12 19 22 Montana 64 36 21 36 33

Appendix III: State Agencies' Progress toward Meeting CMS Survey Goals

Percentage

1998 1999 2000 2001 2002
Nebraska 10 11 25 36 33
Nevada 20 0 8 38
New Hampshire 11 22 0 40
New Jersey 8 12 15 34
New Mexico 16 0 29 14
New Yorka 4 6 10 33
North Carolina a 16 15 21 23
North Dakota 21 31 31 46
Ohioa 13 11 17 38
Oklahoma 13 22 16 41
Oregon 22 65 19 33
Pennsylvania a 5 10 10 11
Rhode Island 10 0 0 15
South Carolina 18 16 17 32
South Dakota 13 19 18 42

Tennessee 11 9 23 47

Texasa 2 7 3 2134 Utah 3530 30 1933 Vermont 00 25 0 33 0 Virginiaa 31 13
12 12 13 Washington 26 18 42 33 32 West Virginia 14 17 13 33 26 Wisconsin
10 16 21 28 30 Wyoming 14 0 0 33 22

Source: GAO analysis of CMS OSCAR data.

aIndicates the 13 states with the greatest number of dialysis facilities
in 2002.

Starting in fiscal year 2001, CMS also set a goal for states to survey all
ESRD facilities every 3 fiscal years. The initial 3-year cycle will be
completed at the end of fiscal year 2003. Table 6 shows the number of
facilities available for recertification in each state at the start of
fiscal year 2001 (and not terminated since then) and the percentage that
remained to be surveyed in fiscal year 2003. In fiscal year 2003, 35 out
of 50 states, plus the District of Columbia, need to survey over a third
of their ESRD facilities to meet the cycle goal.

Appendix III: State Agencies' Progress toward Meeting CMS Survey Goals

Table 6: Facilities to Be Recertified to Meet CMS 3-Year Goal, by State

ESRD facilities needing Share of facilities that need
recertification in fiscal to be surveyed in fiscal
years 2001 through 2003 year 2003 to meet CMS goal
State (number) (percentage)

Alabama 92

Rhode Island 13

California 347

New Mexico 29

Virginia 119

Idaho 7

Michigan 104

Vermont 6

Colorado 35

Missouri 102

Pennsylvania 216

District of Columbia 23

Hawaii 16

Utah 20

Arizona 78

North Carolina 111

Page 47 GAO-04-63 Dialysis Facility Compliance
Maryland 95 47 Iowa 47 47 Delaware 13 46 Texas 285 45 Wyoming 9 44 Nevada 16 44 Wisconsin 76 43 New 87 43 Maine 12 42 Oklahoma 57 40 New 205 40 West 23 39 Washington 43 37 Indiana 78 37 Minnesota 61 36 Louisiana 115 36 South
Jersey York Virginia Dakota 17 35

Appendix III: State Agencies' Progress toward Meeting CMS Survey Goals

ESRD facilities needing Share of facilities that need
recertification in fiscal to be surveyed in fiscal
years 2001 through 2003 year 2003 to meet CMS goal
State (number) (percentage)

34 Oregon 40 35
35 Ohio 149 35
36 South Carolina 75 33
37 Massachusetts 61 33
38 Arkansas 52 29
39 Montana 14 29
40 Nebraska 21 29
41 Illinois 126 26
42 Mississippi 62 26
43 Kansas 39 26
44 Georgia 168 24
45 Tennessee 106 23
46 New Hampshire 9 22
47 Florida 237 22
48 Connecticut 26 19
49 North Dakota 12 17
50 Alaska 2 0

51 Kentucky 45 0

Source: GAO analysis of CMS OSCAR data.

Appendix IV: Comments from the Centers for Medicare & Medicaid Services

Appendix V: GAO Contact and Staff Acknowledgments

GAO Contact Rosamond Katz, (202) 512-7148

Acknowledgments Eric Peterson, Joel Hamilton, Loren Lieberman, and Behn
Kelly made major contributions to this report.

Related GAO Products

Nursing HomeQuality:Prevalenceof Serious Problems, While Declining,
Reinforces Importance of EnhancedOversight. GAO-03-561. Washington, D.C.:
July 15, 2003.

Major Management Challenges andProgramRisks: Department of Health and
Human Services. GAO-03-101. Washington, D.C.: January 1, 2003.

Nursing Homes: Public Reporting of Quality Indicators Has Merit, but
National Implementation Is Premature. GAO-03-187. Washington, D.C.:
October 31, 2002.

Medicare HomeHealth Agencies: Weaknesses inFederal andState Oversight Mask
Potential Quality Issue. GAO-02-382. Washington, D.C.: July 19, 2002.

Nursing Homes: Sustained Efforts Are Essential to Realize Potential of the
Quality Initiatives. GAO/HEHS-00-197. Washington, D.C.: September 28,
2000.

Medicare Quality of Care: Oversightof Kidney Dialysis Facilities Needs
Improvement. GAO/HEHS-00-114. Washington, D.C.: June 23, 2000.

GAO's Mission

Obtaining Copies of GAO Reports and Testimony

The General Accounting Office, the audit, evaluation and investigative arm
of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of
the federal government for the American people. GAO examines the use of
public funds; evaluates federal programs and policies; and provides
analyses, recommendations, and other assistance to help Congress make
informed oversight, policy, and funding decisions. GAO's commitment to
good government is reflected in its core values of accountability,
integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO's Web site (www.gao.gov) contains abstracts
and full-text files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as "Today's Reports," on its
Web site daily. The list contains links to the full-text document files.
To have GAO e-mail this list to you every afternoon, go to www.gao.gov and
select "Subscribe to e-mail alerts" under the "Order GAO Products"
heading.

Order by Mail or Phone The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C.
20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm

Waste, and Abuse in E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Jeff Nelligan, Managing Director, [email protected] (202) 512-4800

Public Affairs U.S. General Accounting Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***